[nonlinear]

Does the Baylis-Hillman reaction work with triethylamine? I know DABCO is usually used. I thought that triethylamine is not very nucleophilic a base... but then, why should DABCO be more nucleophilic?

If you can point me to any literature, that would be helpful as well.

Thanks!

[Honclbrif]

The nitrogens of DABCO are at bridgehead positions which keeps them conformationally locked and consequently makes the lone pairs more exposed and therefore more nucleophilic as compared to other tertiary amines.

[nonlinear]

So it wouldn't be very ideal to use triethylamine in this case, then, correct?

Thanks so much!

[g-bones]

The nitrogens of DABCO are at bridgehead positions which keeps them conformationally locked and consequently makes the lone pairs more exposed and therefore more nucleophilic as compared to other tertiary amines.

So I've heard this argument a lot, about the availability of the lone pairs in comparison to triethylamine(it actually came up in my qualifying exam haha), but I would be forced to believe that the lone pairs should be available in both.  Since the lone pairs on triethylamine are typically in an sp3 orbital, the have a certain reactivity associated with that (Since they don't spend much time in the transition state as being in a p-porbital, amidst inversion), however, due the bond angle constraints associated with the 2,2,2-bicycle of DABCO, the lone pairs on the nitrogen have much more p-character, making them more nucleophilic (higher HOMO).  This is what got me past this topic with my committee 

[Arctic-Nation]

Both Honclbrif and g-bones miss the point that in the Baylis-Hillman reaction, the amine catalyst must act as both a nucleophile and a leaving group. DABCO is used not only because it is a good nucleophile, but also because it is a better leaving group than triethylamine (or related amines).

G-bones, I can't believe that the nitrogen atoms in DABCO would have "much more" p-character than the nitrogen atom in Et3N: DABCO isn't that strained. This is corroborated by the fact that quinuclidine barely differs from Et3N in basicity (pKaH 11.0 vs. 10.7), while it is a much stronger nucleophile (it reacts 60 times faster with MeI than Et3N). Thus, this must be an effect of the alkyl groups being out of the way of the nitrogen lone pairs.

[nonlinear]

Thanks for your responses, guys. I'm an undergrad, so I haven't learned much about DABCO. I was given a mechanism problem in which the only amine present in the reaction (with a methyl vinyl ketone) was Et3N in methanol. I didn't think that Et3N would work given its lack of nucleophilicity, so I proposed an alternative mechanism in which Et3N acted only as a base. However, the answer key shows that Et3N should nonetheless function as a catalyst (so, as a nucleophile and then a leaving group)... my question is, is that technically incorrect? I've read that Baylis Hillman works with NEt3 only under conditions of high pressure and with a proline catalyst.

[Arctic-Nation]

No, the Baylis-Hillman reaction really needs a nucleophilic catalyst or you'll never form the enolate of the vinyl carbonyl compound. As far as I know, the reaction will work with almost any neutral nitrogen nucleophile (and a number of phosphines, too), but it will work best when the catalyst is a good nucleophile (and DABCO is a better nucleophile than Et3N despite its lower basicity, because the alkyl groups are out of the way), and a good leaving group (where DABCO is better because of its lower basicity).

[g-bones]

The 2,2,2 Bicycle is strained (13 kcal/mol for DABCO's hydrocarbon counter part).  The interior angles of the carbon framework lock the nitrogen in more of an sp2 type hybridization, making the lone pair higher in p-character.  You are right that it is a good leaving group and a good nucleophile. The argument of pKa for nucleophilicity is often cited, but it is not always a good model.  Especially in amine systems.  Herbert Mayr is a strong supporter of this theory and has shown repeatedly that pKa and nucleophilicities can differ dramatically.  See ACIE, 2007, 46, 6176 and J. Phys. Org. Chem. 2010,  DOI: 10.1002/poc.1707. 
The alkyl groups of Et3N are out of the way, since it sits mostly as the pyrimidalized conformer (except in the transitions state between inversions, which it doesn't spend much time in of course)

[nonlinear]

Err... so in conclusion, does that mean the Baylis-Hillman WILL or WILL NOT work with NEt3?

[Arctic-Nation]

The 2,2,2 Bicycle is strained (13 kcal/mol for DABCO's hydrocarbon counter part).  The interior angles of the carbon framework lock the nitrogen in more of an sp2 type hybridization, making the lone pair higher in p-character.  You are right that it is a good leaving group and a good nucleophile. The argument of pKa for nucleophilicity is often cited, but it is not always a good model.  Especially in amine systems.  Herbert Mayr is a strong supporter of this theory and has shown repeatedly that pKa and nucleophilicities can differ dramatically.  See ACIE, 2007, 46, 6176 and J. Phys. Org. Chem. 2010,  DOI: 10.1002/poc.1707. 
The alkyl groups of Et3N are out of the way, since it sits mostly as the pyrimidalized conformer (except in the transitions state between inversions, which it doesn't spend much time in of course)

Oh, I most certainly agree that there is no linear relationship between pKaH and nucleophilicity. However, basicity is strongly dependent on the exact composition of the hybrid orbitals, while being virtually independent of steric factors (with simple molecules such as DABCO and Et3N, at least). Since the difference in pKaH between Et3N and quinuclidine is very small (a direct comparison with DABCO is not possible because of the extra nitrogen), this means that the increase in p-character is not the main contributing factor to explain the difference in nucleophilicity. When constructing a molecular model, it should also be clear that the nitrogen's lone pair in quinuclidine and DABCO extend further beyond the rim of carbon atoms than the one in Et3N.

Nonlinear: yes, the reaction will work with Et3N, but when following the standard procedure, it will not work as good as DABCO.

[nonlinear]

Would trimethylamine in MeOH be any different in this reaction?